Degenerate five-carbon scrambling in the 7-norbornadienyl cation

Degenerate five-carbon scrambling in the 7-norbornadienyl cation. Ronald K. Lustgarten, Maurice. Brookhart, and Saul. Winstein. J. Am. Chem. Soc. , 19...
0 downloads 0 Views 335KB Size
6350 Table II. Summary of Solvolysis Products from cisand trans-VIII-OAc VIIIOAc cis trans cis trans

Solvent

Temp, "C

80 % M e 2 C 0

75

Products

99.4 % cis-VIII-OH, 0.6 % trans-VIIIOH 80% M e 2 C 0 100 83.4% cis-VIII-OH, 16.6% transVIII-OH MeOH 100 2.9 % cis-VIII-OH, 97.9 % VIII-OMe (98.9% c i s : l . l % trans) MeOH 100 34.7 % trans-VIII-OH, 65.3 % VIIIOMe ( 9 8 . 7 x cis:1.3z trans)

undetailed nmr signals, as well as a sharp signal at T 0.76 for tropylium ion formed in ca. 25 yield. Despite very rapid decomposition, the nmr spectrum3of ion I could be recorded at 77". At this temperature the T 2.54 signal for the "bound" vinyl protons was still sharp with the coupling pattern somewhat collapsed. However, the T 3.90, 4.88, and 6.73 signals for "unbound" vinyl, bridgehead, and bridge protons were broadened, indicating the onset of an interesting degenerate rearrangement. HvOAc

acid-catalyzed isomerization of the ether product during the quench, the observed product composition is in satisfactory agreement with that observed from kinetic control in acetate methanolysis. While the preference for cis stereospecificity in ionization and formation of covalent VI11 derivatives is probably for stereoelectronic reasons, we are not yet prepared to discuss this in quantum-mechanical terms.' Acknowledgment. We are grateful to Dr. Richard Leute and Mr. George Levy for assistance in some phases of the above research. (7) It is pertinent that NaBHa reduction of VIII-ketone,a mp 84-85", gives 98.8% cis- and 1.2% trans-VIII-OH.

R. F. Childs, S. Winstein Contribution No. 2143, Department of Chemistry Unicersity of California, Los Angeles, California 90024 Received September 5, 1967

Degenerate Five-Carbon Scrambling in the 7-Norbornadienyl Cation' Sir: In this communication we describe a novel degenerate rearrangement of the norbornadienyl cation I discovered during a search for "bridge flipping" by this species2 (Ia e I1 $ Ib). H 76.73 I

-p, / /

,\,e

72.54

\\'' I1

74.88 Ia

Ib

10.76

Extraction of 7-norbornadienol from pentane or CHzClz into FS0,H at -78" gave an F S 0 3 H solution of the 7-norbornadienyl cation I. Warming the solution to +45" caused no noticeable broadening of the nmr signals.2 The ion decomposes at this temperature, as evidenced by the development ot some broad, (1) Research supported by the National Science Foundation. (2) M. Brookhart, R. K. Lustgarten, and S. Winstein, J . Am. Chem. Soc., 89, 6352 (1967).

I11

IVa-OAc

HVOMe

IVb-OH

IVC OMe

To study this rearrangement on a conventional rather than nmr time scale, labeled 7-norbornadienyl precursors were employed. These were IVa-OAc with a syn-vinyl deuterium obtained from acetolysis of 7deuterioquadricyclyl tosylate (111) uia the Richey-Story rearrangement ;4 IVb-OH with a 7-deuterium obtained from 7-deuterioquadricyclanol; and IVc-OMe with 77% of 4-vinyl deuterium atoms obtained by basecatalyzed exchange6 of 7-norbornadienyl methyl ether with LiNDCGHllin C6HllND2. In F S 0 3 H at -73", IVa-OAc displayed the foursignal spectrum of I with the intensity of the unbound vinyl signal only half as great as that for the bound vinyl (Ia). When the cation solution was warmed to - 47 ",a scrambling of the deuterium label was observed with a rate constant of 3 X sec-' (Table I). However, the vinyl proton peak intensities did not approach the 1 : l ratio expected for a bridge-flip phenomenon. Rather, the peak intensities approached a 2 :1.6 : 1.6 :0.8 ratio for the bound vinyl, unbound vinyl, bridgehead, and bridge protons, respectively (Ie). In other words, deuterium was scrambled to all positions except the bound vinyl! The same phenomenon was demonstrated with cations Ib and IC from IVb and IVc, respectively (Table I). At -60" initially, both Ib and IC exhibited the same relative peak intensities as their precursors, but, on warming to ca. -5O", these intensities approached the ratios expected for a five-carbon scrambling reaction (If and Ig, Table I). Examination of the nmr signal intensities during scrambling of Ia, Ib, and IC revealed that deuterium is incorporated sequentially at the different positions. Thus, in the case of Ia and IC, deuterium is first in(3) All chemical shifts are based on CHZCIZas a secondary standard ( r 4.70). (4) (a) H. G. Richey and N. C. Buckley, J . Am. Chem. Soc., 85, 3057 (1963); (b) P. R. Story and S. R. Fahrenholtz, ibid., 86, 527 (1964); 88,374 (1966). (5) A. Streitwieser, Jr., and W. C. Langworthy, ibid., 85, 1757 (1963);

A. Streitwieser, Jr., and J. H. Hammons, Progr. Phys. Org. Chem., 3, 73 (1965). We are grateful to C. Watts for assistance.

Journal of the American Chemical Society / 89:24 / November 22, 1967

6351 Table I. Rates and Free Energies of Activation of Degenerate Isomerizations of the 7-Norbornadienyl Cation

D

R

H

Ib

Ia

Va

Ia

IC

IC

Product

Ie

If

Ig

Temp, "C

-47

-50

i o 4 / ,~ C - L '

3'

-55 1.7'

Al;t,kcal mole

16.7

16.4

Id

Vb

2" 16.7

a Evaluated graphically using In [(Ha - Ho)/(Ha - H I ) ] = kt, with Ha equal to the equilibrium amount of protium. & I nthe unbound vinyl group. At C,.

corporated into the bridgehead, then into the bridge position. With Ib, deuterium appears first at the bridgehead, then in the unbound vinyl group. Thus, the isomerization may be said t o occur by a stepwise circumambulatory motion of five carbons of the framework of I with respect t o the two which comprise the bound vinyl function. This description is supported by the relative line broadenings observed in the hightemperature nmr spectrum.6 The simplest specific mechanism which we can conceive for the 5-carbon degenerate rearrangement involves the ring contraction of I to a bicyclo[3.2.0]heptadienyl cation V by means of a 1,2 shift of a bound vinyl carbon atom from Cl or C4to C7 uia a transition state such as IV, followed immediately by a ring expansion which leaves the same vinyl function bound to

When either cis-VI-OH or trans-VI-OH was extracted into F S 0 3 H at -78", and the carbonium ion solution was observed at - 78 O within 120 sec, the nmr spectrum of I was obtained with no trace of a signal for a [3.2.0] cation V. Most importantly, the a-deuterio cis and trans [3.2.0] alcohols both gave I with 2.0: 1.5 : 1.5 :1.O ratios of signal intensities for bound vinyl, unbound vinyl, bridgehead, and bridge protons, respectively. In other words, 50% of the deuterium label appears at a bridgehead, and 50% in the unbound vinyl group. This result is in exact accord with a mechanism wherein the first formed [3.2.0] cation ring-expands t o the [2.2.1] ion with either carbon of the cyclobutene vinyl function undergoing a 1,2 shift as the migrating vinyl group becomes the bound vinyl of I. The evidence is thus strongly in favor of the ring contraction-ring expansion mechanism for the degenerate five-carbon scrambling of 1 3 9

trcrm-VI-OH

cis-VI-OH

V

H I

V

4 r

l*

H

D'

I

As regards the equilibrium between the [3.2.0] ion V and the [2.2.1] ion I, the latter is strongly favored, A IV

e Va zICe Vb

Id, etc., which shows the bound vinyl group working its way around the five-membered ring defined by the five other carbon atoms in I. This mechanism is supported by the behavior of the cis- and trans-bicyclo[3.2.O]heptadienols'" cis-VI-OH and trans-VI-OH in FS0,H. C7. This is pictured by Ia

(6) The sequential rearrangement requires that the lifetime of the unbound vinyl protons be twice that of the C7 and bridgehead protons and, indeed, the C7 and bridgehead signals are broadened to twice the extent of the unbound vinyl protons. Based on line-width measurements, the rearrangement rate constant at +67" is ca. 40 sec-1. From this rate constant and those in Table I, A H can be estimated as 14.5 f 2.0 kcal/mole.

*

(7) (a) P. R. Story and S. R. Fahrenholtz, J . Am. Chem. SOC.,87, 1623 (1965); (b) P. R. Story, private communication of unpublished work. (8) It is interesting that neither the ring contraction of I to V nor the ring expansion of V to I, which occur in the degenerate five-carbon scrambling of I in FSOJH, where cations are very long-lived, compete with product formation in typical solvolyzing solvents where carbonium ion lifetime is very short. Thus, for example, ion Ia goes cleanly to IVa-OAc in A c O H . ~ Also, ~ according to Story,'b both cis-VI-C1 and trans-VI-C1 give only the [3.2.0]-type product in hydrolysis in aqueous acetone. On the other hand, the Richey-Story rearrangement of the 7-quadricyclyl ion to norbornadienyl (I) does compete with product formation quite well in AcOH4 and even better in HCOOH. (9) The [3.2.0] ion V, which appears to be an intermediate for the five-carbon scrambling of I, could conceivably be an intermediate for the tropylium ion formation. However, other mechanisms are conceivable as well, and this point is still far from clear.

Communications to the Editor

6352

minimum figure for the rate constant for V + I is le2 sec- at 78 ”, and the rate constant for I + V at this sec-l from temperature may be estimated as 1.5 X the available data. Thus, a minimum figure for the equilibrium constant for V e I is 7000 at -78”.

-

(10) (a) National Institutes of Health Postdoctoral Fellow, 19661967; (b) National Institutes of Health Predoctoral Fellow, 1965-1967.

R. K. Lustgarten,lO* M. Brookhart,IobS. Winstein Contribution No. 2158, Department of Chemistry University of California, Los Angeles, California 90024 Received September 18, 1967

Bridge Flipping and Rearrangement of Norbornadienyl and 7-Methylnorbornadienyl Cations’ Sir: The 7-norbornadienyl cation from anchimerically assisted ionization of 7-norbornadienyl derivatives was first studied solvolytically by Winstein and Ordronn e a a 2 Of the possible nonclassical structures visualized by them, the unsymmetrical one represented by I 3 can be chosen on the basis of the nmr spectrum of the ion observed by Story and S a ~ n d e r s , A ~ ~symmetrical structure represented by TI3 was also envisioned, and this is clearly less stable than I. Even at + 4 5 O , the vinyl proton signals for the “bound” and “unbound” vinyl groups in I are harp,^^,^ thus showing that there is a substantial barrier to “bridge flipping” (Ia S Ib) via symmetrical I1 as a transition state (>17 kcal/mole). On the other hand, extended Hiickel calculations6 predict a relatively low value for this barrier (8 kcal; 0.35 ev). The magnitude of this barrier is one of the H I

I1 f/

P

A F t 2 19.6

H T 6.73

subjects of this communication. Methyl substituent effects on the size of the barrier, and on nmr spectra as well, can be quite illuminating, and these are illustrated by means of the 7- and 2-methylnorbornadienyl cations. It was possible t o study the equilibration of bound and unbound vinyl groups in I by observing the labeled cation ICderived from the precursor 111-OMe as a result of degenerate five-carbon scramb1ing.s Cation IChas a deuterium atom at each bound vinyl position but only of a deuterium atom at each unbound vinyl position. At ca. 0” protium incorporation into the bound vinyl positions is indeed observed, the bound and unbound vinyl signals approaching equal intensities. At - 2.5 ”, k = 8 X le4 sec-’ and thus AF* is 19.6 kcal/mole, some 3 kcal greater than AF* for five-carbon scrambling.5

vohfe I\

IC

H

Id

The value of 19.6 kcal/mole is actually a lower limit to the free-energy difference between the symmetrical (11) and unsymmetrical (I) structures of the 7-norbornadienyl cation. This is because bridge flipping is not the only possible mechanism for exchange of vinyl groups. For example, another plausible mechanism involves 1,2 shifts of an unbound vinyl carbon atom7 from C1 t o Cz, or C4 to C3. Such shifts, e.g., I + Ie, move the bound vinyl group around so as to scramble CI, Cz, Ca, C4, and C,. Coupled with the more rapid degenerate five-carbon rearrangement reported prev i o ~ s l yall , ~ seven carbon atoms of I become scrambled.*

I

7

4.88 Ia

Ib

(1) Research supported by the National Science Foundation. (2) S. Winstein and C. Ordronneau, J. Am. Chem. SOC.,82, 2084 (1960). (3) These representations are employed for simplicity. Molecular orbital calculations suggest greater electron delocalization than is indicated: (a) R. J. Piccolini and S . Winstein, Tefrahedron Suppl., 2, 423 (1963); (b) W. G . Woods, Ph.D. Thesis, California Institute of Technology, 1957. (4) (a) P. R. Story and M. Saunders, J, Am. Chem. Soc., 84, 4876 (1962); P. R. Story, ef al., ibid., 85, 3630 (1963); (b) H. G. Richey, Jr., and R.K. Lustgarten, ibid., 88, 3136 (1966). have reported observation of I in FSOaH. ( 5 ) R. K. Lustgarten, M. Brookhart, and S . Winstein, ibid., 89, 6350 (1967). (6) R. Hoffmann, ibid., 86, 1259 (1964).

Journal of the American Chemical Society

As the precursor to the 7-methyl cation V we employed alcohol IV. This was prepared by addition of methyl Grignard to quadricyclanone and subsequent isomerization of the 7-methylquadricyclanol to bicyclic IV with a correct C and H analysis and an appropriate nmr spectrum in CSz. Extraction of I V from CD& into FSOaH at -78” yielded an FS03H solution of V (7) Any involvement of the “unbound” vinyl group in the electron delocalization of I3 would provide some predisposition to such a shift. (8) Still another conceivable mechanism involves the kind of ring contraction occurring in the five-carbon scrambling5 but, instead of the bound vinyl migrating, it is now the less favored unbound vinyl. Ring contraction involving the unbound vinyl group, followed by the conventional ring expansion, results in an exchange of roles between bound and unbound vinyl groups.

1 89.24 1 Nooember 22, 1967